ac ce pt ed articl e · however, 1 equivalent of malonic acid afforded a 40% yield, a 44% drop when...

Supported by

A Journal of

Accepted Article

Title: Gaining Insight Into Reactivity Differences Between Malonic AcidHalf Thioester (MAHT) and Malonic Acid Half Oxyesters (MAHO)

Authors: Sean P. Bew, richard stephenson, jacques rouden, jeremygodemert, haseena seylani, and Luis A Martinez-lozano

This manuscript has been accepted after peer review and appears as anAccepted Article online prior to editing, proofing, and formal publicationof the final Version of Record (VoR). This work is currently citable byusing the Digital Object Identifier (DOI) given below. The VoR will bepublished online in Early View as soon as possible and may be differentto this Accepted Article as a result of editing. Readers should obtainthe VoR from the journal website shown below when it is publishedto ensure accuracy of information. The authors are responsible for thecontent of this Accepted Article.

To be cited as: Chem. Eur. J. 10.1002/chem.201605148

Link to VoR: http://dx.doi.org/10.1002/chem.201605148

FULL PAPER

Gaining Insight Into Reactivity Differences Between Malonic Acid Half Thioester (MAHT) and Malonic Acid Half Oxyesters (MAHO)

Sean P. Bew,[a]* G. Richard Stephenson,[a]* Jacques Rouden,[b] Jeremey Godemert,[a] Haseena Seylani[a] and Luis A. Martinez-Lozano[a]

Abstract: An efficient two-step synthesis of structure- and function-diverse thiophenol- and (cyclo)alkyl-derived malonic acid half thioesters (MAHTs) and phenol-derived malonic acid half oxyesters (MAHOs) has been achieved using cheap, readily available and easily handled starting materials. The synthesis of the MAHTs and MAHOs (majority of which have not been previously reported) is readily scalable affording gram quantities of product. In a hydrogen ® deuterium exchange, an interesting stereoelectronic effect was observed when different aryl groups were incorporated. Significant changes in the rates of hydrogen ® deuterium exchange and levels of isotope incorporation were observed. By way of example, using [2H]methanol and 4-bromophenol-derived MAHO afforded only 14% [2H]-incorporation (9 minutes, k = 31) whereas the corresponding 4-bromothiophenol-derived MAHT afforded 97% [2H]-incorporation (9 minutes, k = 208). In a benchmarked procedure and comprehensive DFT study 54 ester and thioester configurations and conformations where characterised. This established in the MAHT series a sulfur-containing molecular orbital provides a path for increased delocalisation of electron-density into the enol which is unavailable in MAHOs, which facilitates keto-enol tautomerisation and consequently enhances the rate and percentage of hydrogen ® deuterium exchange.

Introduction

Malonic acid half thioesters (MAHTs) and oxyesters (MAHOs) are frequently employed as key starting materials in C-C bond forming reactions e.g. Claisen couplings,[1] aldol[2], Mannich[3] and Michael[4] reactions. The utility of MAHTs (1) and MAHOs (2, Scheme 1) in Nature and the ‘reaction flask’ resides in the reactivity associated with the b-carbonyl carboxylic acid group (red bonds in 1 and 2) and its reaction with different electrophiles. Typical examples include aldehydes, N-tosylimines, trans-b-nitrostyrenes, azodicarboxylates and ketones. The reactions

proceed via similar but ‘inverted’ mechanistic routes either ‘addition-decarboxylation’ pathway i.e. A (Scheme 1) or a ‘decarboxylation-addition’ pathway i.e. B. Depending on where the MAHO or MAHT reactions take place can influence which of these pathways is followed. For example, reacting a MAHO with a pyruvate ester in a ‘reaction flask’, current mechanistic understanding favors ‘Pathway A’.[ 5 ] In contrast, utilizing a polyketide synthase (PKS) and a MAHT, the commonly accepted mechanism[6] involves decarboxylation, generating a thioester enolate (3, Scheme 1), with subsequent addition to an aldehyde affording a b-hydroxythioester e.g. 4, ‘Pathway B’.

As a consequence of MAHT-enhanced reactivity, there is increasing interest in developing novel activation pathways for MAHTs; organocatalysis (vida infra) and metal-mediated catalysis are leading examples.[7]

Detailed studies on the reactivity differences of MAHTs and MAHOs have not been conducted. We propose these are related to their ease of keto-enol tautomerization. Probing the ability of MAHTs and MAHOs to tautomerization we report here insight into their reactivities using 1H-NMR and a hydrogen ® deuterium exchange protocol using structure- and function-diverse substrates based on 1 and 2. A subsequent computational study afforded a better understanding of how changing the substituents on the aryl group in sulfur-stabilized MAHTs e.g. cisoid-5 (Z = S, Scheme 1) and enol-stabilized MAHOs e.g. cisoid-6 (Z = O) can enhance or decrease their reactivity.

Scheme 1. ‘Addition-decarboxylation’ and ‘decarboxylation-addition’ reaction pathways associated with MAHTs and MAHOs.

PKSs are large multi-domain enzymes that use MAHTs for the biosynthesis of fatty acids and biologically active secondary metabolites.[ 8 ] Examples of some important

[a] Dr S. P. Bew, G. R. Stephenson, J. Godemert, H. Seylani, L. A. Martinez-Lozano School of Chemistry Norwich Research Park

University of East Anglia NR4 7TJ, UK E-mail: [email protected] or [email protected]

[b] Dr J. Rouden Laboratoire de Chimie Moleculaire et thio-organique (LCMT) UMR CNRS 6507, Ensicaen 6 Boulevard du Marechal Juin, 14050 Caen, France.

Supporting information for this article is given via a link at the end of the document.

S

O

OH

OO

O

OH

OAr Ar

RX

O

OH

O

Ar' H

O

Ar' H

O

X = O, S+ RX

O

O-

O

Ar' OH

MAHT, 1 MAHO, 2

- CO2

enzyme

RS

O

Ar'

OH

RS

O-+

- CO2X = S

thioester enolate

β-carbonyl carboxylic acid

Addition-decarboxylation,Pathway A

Decarboxylation-addition,Pathway B

Z

O

OH

OAr

Z

O

OH

OAr

HX

Z

O

OH

OAr

HX

Z = S Z = O

MORE reactive MAHT LESS reactive MAHO

cisoid-5 (X = Br) cisoid-6 (X = Br)

3

4

10.1002/chem.201605148Chemistry - A European Journal

This article is protected by copyright. All rights reserved.

FULL PAPER

secondary metabolites include rifamycin (antibiotic), callystatin (anticancer), psymberin (anticancer), pridamicin (antifungal) and 6-deoxyerythrolide (antiviral). ‘Activation’ of the MAHTs in PKS is via an active site embedded histidine (His) and asparagine derived CONH2 (Asn, Scheme 2). In this metal-free environment the CoA thioester malonic carboxylate anion is hydrogen-bonded (7) to protonated His and Asn. The activated MAHT eliminates CO2 and generates reactive thioester enolate 8. Due to its reactivity and close proximity to an electrophilic cysteine thioester it affords CoA bound b-keto thioester 10 via tetrahedral intermediate 9.[9]

Efforts to build natural or artificial PKS constructs that mimic the capabilities of CoA-derived malonic thioesters cf 7 are held back by a lack of understanding of the reactivity, conformation and electronic ‘communication’ within 1 and 2. Thus developing efficient, mild and sustainable protocols for structure- and function-diverse MAHO(T)s is important in their applications in biomimetic and synthetic processes.

Scheme 2. Biosynthetic acetyl-CoA derived MAHT ‘activation’, thioenolate formation and aldol condensation to generate acetyl-CoA bound b-ketothioester.

Increasing interest in the synthesis applications of MAHO(T)s can be attributed, in part, to their exploitation in asymmetric organocatalysis.[ 10 ] However, there are reports alluding to the difficulty in generating oxyester enolates of 1 that attribute this to the high pKa values of phenoxy esters i.e. ~18 (DMSO).[ 11 ] Therefore, increased emphasis has been placed on phenylthioesters which, generally, have lower pKa’s ~16-17 (DMSO).[ 12 ] Substituting MAHTs for MAHOs, their application in organocatalytic protocols becomes viable. By way of example, Song et al. utilized 12 (Scheme 3) in conjunction with squaramide-derived organocatalyst 14 for the enantioselective Michael addition to trans-b-nitrostyrene 11 generating g-amino acid precursors, based on 13 (88 - 99% e.e. and 15 - 96 yields).[ 13 ] Similarly, using cinchinone-derived catalyst 18 conjugate addition of 15 to coumarin 16 afforded thioester 17 in 95% - 99% yields and 84% - 92% e.e.’s.[14] Finally, Ricci et al. synthesised 20 via an enantioselective ‘addition-decarboxylation’ of aryl- or alkylthio-MAHTs to 19 using cinchona-derived catalyst 21. The b-N-tosylamino

thioesters (20) were produced in 21-79% e.e. and 42 - 84% yields.[15]

Scheme 3. Bae et al. Enantioselective biomimetic Michael addition of MAHT 12 to trans-b-nitrostyrene catalyzed by cinchona-based 14. Nakamura et al. Enantioselective Michael addition reaction catalyzed by 18 generating thioester 17. Ricci et al. Synthesis of b-N-tosylamino thioester 20 via ‘addition-decarboxylative’ reaction of a MAHT to 19 catalyzed by 21.

Increasing applications of MAHOs and MAHTs reinforces the need for straightforward, efficient and scalable syntheses. Furthermore, there are relatively few syntheses of structure- and function-diverse aryloxy-derived MAHOs and arylthiol-derived MAHTs. The majority of MAHOs are generated using 2,2-dimethyl-1,3-dioxane-4,6-dione (22).[ 16 ] However, its thermal instability to elevated temperatures, requirement for cold storage and, unless freshly bought, recrystallization before use.[17] Kee et al. reacted phenol with 22 generating MAHO 23 (Scheme 4).[18] Mase et al. generated 25 (62% yield) from malonic acid and 24 using dicyclohexylcarbodiimide (DCC) as the coupling agent.[ 19 ] However, DCC is not particularly atom-efficient and although not commented on, removing trace amounts of dicyclohexylurea (DCU) from reaction products can be problematic.

Searching SciFinder for aryl thioester MAHTs generated 32 references, these contained 10 unique MAHTs. Further analysis revealed, rather surprisingly, only 2 different synthesis protocols had been utilized. Using Meldrum’s acid, Kee et al. generated 15 in a 24% yield,[16] unfortunately it also generated difficult to remove dithiophenylmalonic ester by-product (not shown) afforded in a near equal percentage i.e. 21%. Imamoto et al. reported the synthesis of 8 MAHTs. The series comprised 5 alkyl and the 3 aryl thioesters e.g. 15, 29 and 3-(2,4,6-trimethylphenylthio)-3-oxopropanoic acid (not shown) generated in 88%, 65% and 88% yields respectively. Potential problems and limitations reside in the use of 4 equivalents of malonic acid and not commercially available ethyl polyphosphate (also used in excess), and the long reaction time (30 hours). Imamoto probed the requirement for excess



FULL PAPER

malonic acid during the synthesis of 15. However, 1 equivalent of malonic acid afforded a 40% yield, a 44% drop when

compared with 4 equivalents.[20]

Scheme 4. Synthesis of MAHOs 23, 25, 28 and MAHTs 15 and 29.

Results and Discussion

Our strategy had 3 components: 1. Part 1 - Synthesis. Our aim was to develop a simple,

versatile, mild and efficient protocol for generating structure and function diverse thioaryl- and thioalkyl-derived MAHTs and aryl oxyester-derived MAHOs. It was important they were readily available from cheap ‘off the shelf’ reagents using an operationally straightforward protocol.[ 21 ] Furthermore, it was essential our protocol was applicable to the synthesis of sterically encumbered, electron-rich or electron-poor MAHO(T)s.

2. Part 2 - Reactivity. Exploring MAHO (1) and MAHT (2) reactivity, a comprehensive 1H-NMR structure activity

relationship (SAR) study sought to exploit their acidic a-CH2’s and probe the roles of the sulfur / oxygen in their ability to effect keto-enol tautomerisation in a hydrogen → deuterium exchange.

3. Part 3 - Computation. DFT studies were deployed to explain / determine the unusual reactivity displayed by some MAHTs in the hydrogen ® deuterium exchange study (Part 2). Crucial differences in the delocalised molecular orbitals associated with the enol were identified and may account for their differing reactivities.

Initiating our work we developed a straightforward synthesis of malonyl mono-chloride[22] using readily available malonic acid and thionyl chloride. Stirring a mixture of the starting materials for 6 hours in ether at reflux, followed by removal of any volatile compounds (high vacuum), afforded a product suitable for use ‘as is’ in onward reactions. Without further purification 50 g batches of malonyl mono-chloride were readily synthesized and stored at ambient temperature, for at least 3 months, with no signs of decomposition.

Our next objective was to generate aryloxy-derived MAHOs. Currently, within the synthetic chemistry community, there is considerable interest in the development of solvent-free protocols.[23 ] Attempting the synthesis of 23 (Table 1), malonyl mono-chloride and phenol were mixed at ambient temperature. After a short period of time a slow coalescence towards a homogenous solution was observed which, if left for 2 hours, afforded complete dissolution. Stirring the reaction for a further 4 hours, a small aliquot was removed for analysis after dissolving in [2H]chloroform. Disappointingly, 1H-NMR (400 MHz) indicated the majority of 30 (Scheme 5) was still present and only a small amount (~10%) of MAHO 23 had formed. Although it seemed this approach had merit, the reaction was incomplete and the time prohibitively long. The most straightforward way to increase the rate and hopefully generate 23 more efficiently was to raise the reaction temperature. At 100 °C the mixture quickly transformed into a homogeneous clear pale yellow solution. After 30-minutes it was cooled to room temperature and dichloromethane (DCM) added. The white precipitate was filtered, dried under vacuum and a small portion re-dissolved in [2H]chloroform. Gratifyingly 1H-NMR analysis confirmed 23 had formed, with minimal decomposition, in an unoptimized 75% yield.

Malonic acid derivatives are, generally, thermally unstable. For subsequent studies it was important to determine the rate of reaction between phenol and malonyl mono-chloride (no solvent) at 100 °C. Probing the synthesis of 23 it was analyzed at 4-, 9- and 19-minute intervals. 1H-NMR spectrometry indicated consumption of phenol and, in all cases, efficient formation of 23. The reaction was repeated, again without solvent, in a preheated bath (100 °C) and rapidly cooled (ice-bath) after 2 minutes. 23 was isolated and purified via flash-column chromatography in a 77% yield. Thus, employing these conditions the synthesis of MAHO 23 was fast, efficient and high yielding. As previously noted, it was important our protocol was scalable affording, if desired, multi-gram quantities of MAHO(T)s. 10 g of malonyl mono-chloride was reacted with phenol (100 °C for 2 minutes) followed by

OHHO

O O

HN

OO NH

O

HN

OO NH

O

O

HN

OO NH

O

HO2CHO2C

HO2C3 3

3

1. benzyl malonylchlorideDCM,

pyridine

OH

2. Reaction repeated

32% yield

CH3OHPd / C / H2

95% yield(+ 5 - 10%

of 23)

23 24

25

PhO OH

O O

H3COOH

+ O

O

H3CO

EtOAcDCC

10°C theno/n at rt

62% yield

O O

OO

+OH

90 °C4 hrs

92% yield20 21

22

O OBn

O OH

OHO

HO

HOO

O

PhS OH

O O

OCH3

+ S

O

THF / CHCl330 hrs rt

65% yield

O O

OO

+SH

90 °C4 hrs

24% yield20 13

26

SHOCH3

ethylpolyphosphate

HO O



FULL PAPER

cooling and addition of DCM. Analysis of the precipitate confirmed 23 (13 g) had formed cleanly and efficiently in a 77% yield (Scheme 5).

Scheme 5. Efficient, multi-gram solvent-free synthesis of 23

To facilitate ‘take up’ of our protocol it was important to demonstrate synthetic versatility. We probed the formation of sterically encumbered MAHOs by incorporating hindered 2,6-dimethylphenol, 2-tert-butylphenol as well as the extremely hindered 2,6-di-tert-butylphenol. Reacting 2,6-dimethylphenol and malonyl mono-chloride in our standard reaction conditions (i.e. 100 °C for 2 minutes) afforded after work up a white solid. Subsequent physicochemical analysis confirmed 41 (Table 1) had formed, in an unoptimized 66% yield. Increasing the steric bias further MAHOs derived from 2-tert-butylphenol and 2,6-di-tert-butylphenol were attempted. Using our standard reaction conditions afforded the previously unknown 42, and 32 (Scheme 6). Both were isolated as easily handled, stable solids in 65% and 43% yields respectively. Although 32 was afforded in slightly lower isolated yield than 41 and 42, the reaction seems fully amenable to generating highly sterically encumbered MAHOs. Changing tack, we sought to incorporate a sterically less challenging 2-allyphenol substituent. Employing standard conditions 43 (Table 1) was afforded with an embedded allyl group, a versatile chemical handle, in close proximity to the b-keto carboxylic acid in a pleasing 66% yield.

Scheme 6. Synthesis of sterically encumbered MAHO 32

We turned our attention to the synthesis of MAHOs derived from electron-deficient substrates e.g. 2- and 4-nitrophenol, 2- and 4-cyanophenol as well as methyl 2- and 4-hydroxybenzoate. In the 4-substituted series, MAHOs 44 (4-nitrophenol), 45 (4-cyanophenol) and 47 (methyl 4-hydroxybenzoate) were all readily generated in good 66%, 60% and 51% isolated yields. Seemingly the known propensity of phenols to undergo intermolecular hydrogen-bond dimerization does not appear to hinder the formation of the MAHOs. In contrast, the incorporation of 2-nitrophenol or methyl 2-hydroxybenzoate resulted in poor isolated yields of 37 and 40 i.e. 25% and 21% respectively. Interestingly 2-

cyanophenol afforded MAHO 46 with a restored and considerably higher (than 37 and 40) 61% yield (Table 1).

Table 1. Examples of structure and function diverse MAHOs which are readily generated using a mild, efficient and straightforward reaction protocol.

No Name Yield[a]

23 3-oxo-3-(phenoxy)propanoic acid 77%

25

32

37

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

3-oxo-3-(4-methoxyphenoxy)propanoic acid 3-oxo-3-(2,6-di-tert-butylphenoxy)propanoic acid 3-oxo-3-(2-nitrophenoxy)propanoic acid 3-oxo-3-((2-methoxycarbonyl)phenoxy)propanoic acid 3-oxo-3-(2,6-dimethylphenoxy)propanoic acid 3-oxo-3-(2-tert-butylphenoxy)propanoic acid 3-oxo-3-(2-allylphenoxy)propanoic acid 3-oxo-3-(4-nitrophenoxy)propanoic acid 3-oxo-3-(4-cyanophenoxy)propanoic acid 3-oxo-3-(2-cyanophenoxy)propanoic acid 3-oxo-3-((4-methoxycarbonyl)phenoxy)propanoic acid 3-oxo-3-(4-hydroxyphenoxy)propanoic acid 3-oxo-3-(2-hydoxyphenoxy)propanoic acid 3-oxo-3-(2-methoxyphenoxy)propanoic acid 3-oxo-3-(4-chlorophenoxy)propanoic acid 3-oxo-3-(4-bromophenoxy)propanoic acid 3-oxo-3-(4-iodophenoxy)propanoic acid 3-oxo-3-(2-chlorophenoxy)propanoic acid 3-oxo-3-(2-bromophenoxy)propanoic acid 3-oxo-3-(4-trifluoromethylphenoxy)propanoic acid 3-oxo-3-(3-trifluoromethylphenoxy)propanoic acid

61%

43%

25%

21%

66%

65%

66%

66%

60%

61%

51%

nd

nd

45%

64%

69%

77%

68%

71%

55%

59%

[a]. nd = not determined We sought to rationalize the low yields for 37 and 40.

Here we propose an intramolecular hydrogen-bond in 2-nitrophenol and methyl 2-hydroxybenzoate perturbs, in a negative sense, their reaction with malonyl mono-chloride. A substantial body of evidence supports 2-nitrophenol’s existence in a “closed” i.e. 34 (Scheme 7) or “open” (not shown) conformation. When “closed” the resulting resonance-assisted hydrogen bond (RAHB) generates non-esterifiable o-quinonoid 33. Furthermore, in conformation 34 (“closed”) its reactivity is significantly curtailed by a strong hydrogen-bond, which for 2-nitrophenol has been determined using different experimental techniques to be 25 - 35 kJ mol-1.[24] Interestingly, ab initio studies predict a considerably higher 45 - 50 kJ mol-



FULL PAPER

1.[25] Needless to say which ever is considered, the strength of the hydrogen-bond is reflected in the short ca. 1.7 Å OH····O bond (based on X-ray diffraction[ 26 ]). The net result of the RAHB and the inductive effect (2-nitro group) is a reduction in the nucleophilicity of the phenolic hydroxyl of 34 affording it with poor reactivity towards acyl chloride 30. A similar argument is proposed for 39. Again, a strong intramolecular hydrogen bond i.e. OH····O (< 2.0 Å) combined with RAHB (38) and a strong inductive effect is highly effective at reducing the nucleophilicity of the phenolic hydroxyl group resulting in a low 21% yield of 40 (Table 1).

As a homogenous reaction it was prudent to consider the effect (if any) malonyl mono-chloride may have as a bifunctional hybrid-like and unconventional malonic acid-malonyl chloride ‘solvent’ on the poor yields of 37 and 40. As a potential proton donor, we considered the possibility the ‘malonic acid’ (pKa 2.5) may protonate the phenolic oxygen of 34 generating malonate ion-pair 35 (Scheme 7) on route to 37. However, the phenolic hydroxyl is poorly nucleophilic and as such it will be reluctant to undergo protonation by the weakly acidic carboxylic acid of 30, furthermore once protonated 35 is no longer able to participate in the desired reaction pathway. It seems feasible 30 could protonate “closed” 34 via the pink arrows (Scheme 7) and disrupt the strong intramolecular hydrogen-bonded complex i.e. 34 → 36. However, as a major reaction pathway this is unlikely because it requires breaking the strong OH····O=N(O)- bond. Applying similar RAHB arguments (i.e. formation of 38) and reasoning, the low yield for ortho-ester 40 can be rationalized. Thus, it seems 37 and 40 are generated via an inefficient reaction of poorly nucleophilic 34 or 39 with malonyl mono-chloride i.e. 39 → 40 (Scheme 7).

Scheme 7. Poor reactivity of 34 and 39 is ascribed to RAHB and strong inductive effects associated with the 2-nitro and 2-carboxymethyl groups.

Reacting one equivalent of malonyl mono-chloride with electron-rich hydroquinone or catechol was at first promising. Although the isolation of the presumed mono-carboxylate 48 and 49 (Table 1) together with what appeared to be small

quantities of the corresponding dicarboxylic acids (not shown) was straightforward (filtration), their insolubility in a wide range of solvents made further purification tedious, difficult and ultimately not possible. The synthesis of analytically pure, electron-rich aryl-derived MAHOs was, however, an important milestone. Incorporating the more lipophilic 4-methoxyphenol and 2-methoxyphenol the desired mono-carboxylic acids 25 and 50 were isolated as easily purified solids in 61% and 45% yields respectively (Table 1). Once again the 2-regioisomer (50) was afforded in a slightly lower than anticipated yield (cf. 37 and 40). Presumably, 2-methoxyphenol generates a five-membered intermediate strength intramolecular hydrogen-bond (2.11 Å27), similar to 34 and 39, that reduces its reactivity and results in a lower yield of product (50).

Synthesizing aryl halide derived MAHOs was important because of their applications as easily accessible starting materials capable of metal-catalyzed chemical manipulation and elaboration.[ 28 ] 4-Chloro, 4-bromo and 4-iodophenol reacted, without solvent, with malonyl mono-chloride (100 °C, 2 minutes) affording 51, 52 and 53 in broadly similar 64%, 69% and 77% yields respectively. Furthermore, unlike the incorporation of strong EWGs at the 2-positions of 37 (nitro) and 40 (methyl ester) the 2-chloro- and 2-bromophenol adducts afforded 2-chlorophenyl-54 and 2-bromophenyl-55 in 68% and 71% yields respectively. Explaining the relatively weak ‘ortho-halo effect’ it is known they form only moderately strong intramolecular X····HO bonds (X = Cl or Br).[ 29 ] Incorporating substrates equipped with more than one halogen was explored using the trifluoromethyl group. Both 4-(trifluoromethyl)phenol and 3-(trifluoromethyl)phenol afforded 56 (55% yield) and 57 (59% yield, Table 1). Clearly incorporting the trifluoromethyl moiety is not detrimental to a successful reaction outcome.

Confident our protocol was amenable to generating structure and function diverse aryloxy-derived MAHOs attention switched to the corresponding MAHTs. Utilizing the protocol for generating 23 i.e. neat, 100 °C, 2 minutes but with thiophenol we were surprised 15 was afforded in a poor 25% yield (cf. 77% yield for MAHO 23, Table 1). Presumably the greater reactivity of 15 relative to MAHO 23 resulted in decomposition at the elevated reaction temperature. Attempting to mitigate this the rapid two-minute reaction time was maintained whilst the temperature was lowered to 65 °C.

Disappointingly 15 was afforded in a poor yield with the reaction comprising largely unreacted thiophenol and 30. If, however, it was increased to two hours whilst maintaining the temperature at 65 °C 15 was isolated as a white solid after a simple work-up in a 59% yield. Our slightly modified protocol proved to be a convenient starting point for the incorporation of 16 commercially available thiophenols. Monocyclic 4-methylthiophenol, sterically encumbered 2,6-di-methylthiophenol and multicyclic 2-naphthalenethiol all reacted (individually) with malonyl mono-chloride. The desired MAHTs 58 - 60 were isolated in 59%, 43% and 52% yields respectively with the reaction appearing to be unaffected by steric bias, thus 2,6-dimethylthiophenol and 4-methylthiophenol afforded, within experimental error, identical yields. Similar to the



FULL PAPER

synthesis of electron-poor 4-nitrophenyl derived MAHO 44 (66% yield), 4-nitrothiophenol was readily transformed into the previously unknown 61 in a 52% yield. 2-Nitrothiophenol afforded the expected MAHT i.e. 62 in a 30% yield. Comparing 37 (25% yield) with the slightly higher yield of 62 can be accounted for in terms of the thiol group being more nucleophilic than the hydroxyl and within the 2-nitrothiophenol the intramolecular hydrogen-bond i.e. –NO2····HS-30 (1.922 Å and 34.64 kJ mol-1) being longer and weaker than the hydrogen-bond in 37.

Incorporating electron-rich 4-methoxythiophenol and 2-methoxythiophenol the corresponding 12 and 29 (Scheme 4) were afforded in 61% and 51% yields respectively. The identical yields for 4-methoxyphenyl-MAHO 25 and 4-methoxythiophenyl-MAHT 12 (Table 2) verify the viability of incorporating EDGs on phenols or thiophenols and their suitability as starting materials for novel MAHO(T) synthesis.

Similar to the reasons outlined for aryl halide containing MAHOs 51 - 57 (Table 1), the incorporation of thioaryl halides and (trifluoromethyl)thiophenols in MAHTs was important. Employing our ‘standard’ MAHT reaction conditions malonyl mono-chloride was added, independently, to 4-chlorothiophenol, 3-chlorothiophenol, 2-chlorothiophenol, 4-bromothiophenol, 4-fluorothiophenol and 4-trifluoromethylphenol. The desired MAHTs 63 - 68 were isolated in 50% (4-chlorophenyl), 58% (3-chlorophenyl), 45% (2-chlorophenyl), 41% (4-bromophenyl), 65% (4-fluorophenyl) and 50% (4-trifluorophenyl) yields (Table 2). Reflecting on the slightly lower yield of 65 it is known, similar to 2-chlorophenol, an ortho-chlorine atom can generate a 5-membered intramolecular hydrogen bond (2.16 Å) with the adjacent thiol i.e. -Cl····HS-[31] which reduces the nucleophilicity of the thiol, albeit not to the same extent as a 2-nitro group, and impedes its capacity to react efficiently with 30. Similar to 56 and 57 4-(trifluoromethyl)aryl substituted MAHT 68 was readily generated in a 50% yield and, within experimental error, an identical yield to 4-(trifluoromethyl)aryl substituted MAHO-56 i.e. 55% yield. Disappointingly, all attempts at generating pentachlorophenyl-69 failed with, at best, single figure yields returned. Presumably the five electron-withdrawing chlorine atoms reduce the nucleophilicity of the thiol to such an extent it is no longer able to react with acyl chloride 30. Expanding the substrate scope, we sought to include representative examples of alkyl thiols. 3-Mercaptopropanoate methyl ester, benzyl mercaptan and cyclohexanethiol generated 70, 71 and 72 in 30%, 55% and 44% yields respectively (Table 2).

Ultimately it is our intention to develop novel isotope derived MAHOs and MAHTs as multinuclear NMR[32] and mass spectrometric (MS) probes.[33] Therefore our focus switched to generating structure and electronically diverse isotopologues [2H]-MAHO (73) and [2H]-MAHT (74) (Scheme 8).

MAHOs and MAHTs are important motifs in synthetic and biological chemistry. (Scheme’s 2 and 3). Similarly, stable isotopes, especially deuterium, are widely employed in many different research sectors e.g. pharmaceutical, agrochemical, biotech, medtech, materials and analytical chemistry. However, only a handful of deuterated MAHOs and MAHTs have been

reported; apart from several recent examples from our own laboratory and Matille et al.[21] only two are known i.e. 2-methyl-3-oxo-3-(phenylthio)-[2H]propanoic acid and 2-([2H]methyl)-3-oxo-3-(phenylthio)propanoic acid.[34] Table 2. Examples of structure and function diverse MAHTs which are readily generated using a mild, efficient and straightforward reaction protocol.

No Name Yield[a]

12 3-oxo-3-(4-methoxyphenylthio)propanoic acid 61%

15

29

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

3-oxo-3-(3-phenylthio)propanoic acid

3-oxo-3-(2-methoxyphenythio)propanoic acid

3-oxo-3-(4-methylphenylthio)propanoic acid

3-oxo-3-(2,6-dimethylphenylthio)propanoic acid

3-oxo-3-(2-napththylthio)propanoic acid

3-oxo-3-(4-nitrophenylthio)propanoic acid

3-oxo-3-(2-nitrophenylthio)propanoic acid

3-oxo-3-(4-chlorophenylthio)propanoic acid



3-oxo-3-(4-bromophenylthio)propanoic acid

3-oxo-3-(4-fluorophenylthio)propanoic acid

3-oxo-3-(4-trifluoromethylphenylthio)propanoic acid

3-oxo-3-(pentachlorophenylthio)propanoic acid

3-(2-methoxy-2-oxoethylthio)-3-oxopropanoic acid

3-oxo-3-(benzylthio)propanoic acid

3-oxo-3-(cyclohexylthio)propanoic acid

59%

51%

59%

43%

52%

52%

30%

50%

58%

45%

41%

65%

50%

2%

30%

55%

44%

We contemplated the synthesis of entities based on [2H]-

73 and [2H]-74 using [2H]malonic acid ([2H]-75, Scheme 8). We had reservations however on progressing deuterated intermediates through multiple procedures and, potentially, drawn-out flash chromatography purifications using ‘wet’ and acidic silica gel. With our concerns focused on the product being produced with reduced levels of [2H]-incorporation resulting from deuterium → hydrogen exchange (‘washout’) at the acidic [2H]-methylenes. Opting for an alternative strategy we took advantage of the facile keto-enol tautomerisation properties of ‘pre-assembled’ MAHO(T)s and their potential for hydrogen ® deuterium exchange. At the outset, it was important to establish a proof of principal using a simple MAHO and MAHT. Ester 23 and thioester 15 were dissolved in [2H]methanol and monitored by 1H-NMR at 4-, 9-, 19- and 29-minutes. Table 3 outlines the rather surprising results with significantly divergent levels of [2H]-incorporation for these very similar compounds.



FULL PAPER

Scheme 8. Potential synthesis of MAHT [2H]-77 from [2H]-75

Evidently after only 4 minutes the extent of deuterium exchange for thiophenol-derived MAHT 15 was significantly faster and more efficient than the corresponding phenol-derived 23. Compare, for example, after only 4 minutes 72% incorporation for 15 with only 18% for 23. Furthermore, after only 9 minutes 15 had incorporated 90% deuterium, 52% more than its chalcogen equivalent. After 29 minutes 23 still had not undergone complete isotope incorporation reaching a level equivalent to the same afforded by 15 in only 4 minutes! Clearly, switching phenol for thiophenol has a significant and positive effect on the enol reactivity of 15. It was important to ascertain if the increased reactivity was an ‘isolated’ result specific to 15 or if it was a more general and if so more interesting, presumed, stereoelectronic effect. Using a selection of structure and function diverse MAHOs (17 examples) and MAHTs (7 examples) a comprehensive hydrogen ® deuterium exchange study was undertaken. Employing a standard concentration of MAHO or MAHT (both 70 µM) in excess [2H]methanol (500 µL) the results are presented in Figure 1. What is immediately evident is high-levels of [2H]-incorporation can be readily achieved in the majority of substrates simply by stirring them at ambient temperature with a deuterated protic solvent. Furthermore, there are significant rate differences between thioesters and esters [compare graphs (a) and (b)] but also within each of the two classes different aryl substituents have a significant influence on their reactivity and ability to undergo hydrogen ® deuterium exchange. See, for example, the clear ‘gap’ between the majority of examples and the four electron-poor arenes (Figure 1). Our observation is further reinforced by comparing 4-bromothiophenol-derived MAHO 52 with 4-cyanophenol-derived MAHO 45. Not only is there a noticeable difference in the rate of exchange but also a significant difference in the percentage of [2H]-incorporation. Compare 2,6-dimethylphenol-derived 59 which has nearly 50% [2H]-incorporation after only 9 minutes with 4-bromophenol-derived 52 which is less than 10%. Further highlighting the reactivity differences between almost identical

MAHOs and MAHTs is the surprising observation with 4-bromothiophenol-derived 66 which afforded the largest percentage of [2H]-incorporation, after only 9 minutes (96%). This contrasts sharply with 4-bromophenol-derived 52 which, as already noted, afforded only 15% [2H]-incorporation - the only difference between them, a sulfur versus an oxygen atom. Table 3. Percentage levels of [2H]-incorporation into MAHO 23 and MAHT 15 after 4-, 9-, 19- and 29-minute reaction times.

Time Phenol MAHO 23

4 18% [2H]-incorporation

9

19

29

Time

4

9

19

29

38% [2H]-incorporation



Thiophenol MAHO 15





Dissecting the data generated from the hydrogen ®

deuterium study (Figure 1) several important observations and interpretations can be inferred:

1. Using easy to generate 1H-NMR data on MAHOs and MAHTs it is possible to determine fundamental reaction rate differences between these two classes of compounds.

2. Integrating the remaining protons on the acidic a-methylenes of the MAHOs and MAHTs after 9 minutes afforded a convenient ‘snap-shot’ of the initial rate of hydrogen ® deuterium exchange. Here we determined the percentage of [2H]-incorporation and the relative rates of reaction (k) associated with the exchange process.

3. The results (Table 4) clearly show MAHOs derived with 2- and 4-bromophenol undergo hydrogen ® deuterium exchange at approximately the same rates i.e. k = ~31 ±2 and afford almost identical low levels of [2H]-incorporation i.e. 14% - 15%. In the same study, 4-methoxyphenol-derived 25 (k = 60), 4-trifluoromethylphenol-derived 56 (k = 67) and 4-chlorophenol derived 51 (k = 66) afforded, at similar rates of reaction, the [2H]-MAHOs with 27%, 30% and 31% [2H]-incorporation. Clearly, hydrogen ® deuterium exchange in these MAHOs is faster and more efficient than the bromoaryl-substituted MAHOs.

4. Strong electron-withdrawing para- and ortho-substituted aryl groups on MAHOs increase (see Table 4) the rate of hydrogen ® deuterium exchange, see for example 44 (4-nitrophenyl), 37 (2-nitrophenyl)



FULL PAPER

and 47 (4-carboxyphenyl methyl ester). For example, 4-cyanophenol-derived 45 affords 79% [2H]-incorporation with a rate k = 171 whereas phenol-

derived 23 affords 38% [2H]-incorporation with a reduced rate i.e. k = 83.

Figure 1. Rates of hydrogen ® deuterium exchange for a variety of structure and function diverse MAHO and MAHOs.

5. Within the confines of our study the overall rate of

hydrogen ® deuterium exchange is significantly faster in the majority of MAHTs than MAHOs.

6. Increasing the steric bulk either side of the phenolic ester alcohol increases the rate of hydrogen ® deuterium exchange but with only moderate increases (after 9 minutes) of [2H]-incorporation. Compare for example phenol-derived 23 (k = 83) with 2,6-dimethylphenol-derived 41 (k = 101).

7. In like-for-like comparisons incorporating specific ortho-substituents a decrease in the rate and percentage of [2H]-incorporation with respect a para-substituent was observed. For example, methyl ortho-ester-derived MAHO 40 returned 24% [2H]-incorporation (k = 53) and ortho-cyano-derived MAHO 46 38% [2H]-incorporation (k = 85). Whilst the analogous para-methyl ester-derived 47 returned 72% [2H]-incorporation (k = 157) and para-cyano- derived 45 afforded 79% [2H]-incorporation (k = 171). Similarly, ortho-methoxythiophenol-derived 29 afforded lower [2H]-incorporation (46%) and at a slower rate (k = 102) than para-methoxythiophenol 12 with 74% [2H]-incorporation (k = 161). Differences in the levels of [2H]-incorporation in between ortho-

and para-substituted MAHO(T)s seem dependent on steric congestion at the ortho-substituents.

8. The distributions of the hydrogen ® deuterium exchange reaction rates are smaller, within the 9-minute window (see Table 4), for the MAHTs i.e. k = 102 ® 208 than for the MAHOs i.e. k = 30 ® 171. These results suggest, in general, the rate of exchange for MAHTs is less dependent on the aryl substitution than the MAHOs.

9. In nearly all examples the initial rate of hydrogen ® deuterium exchange for the MAHTs were substantially faster than the majority of MAHOs. The only exceptions were when MAHOs were appended with strong electron-withdrawing groups i.e. 44 (4-nitrophenol), 37 (2-nitrophenol), 45 (4-cyanophenol) and 47 (4-carboxyphenyl methyl ester).

10. All the MAHTs afforded 95% [2H]-incorporation after less than 39 minutes. Indeed, excellent levels of deuterium were installed in many MAHTs in less than 20 minutes. Further highlighting the reactivity differences between MAHTs and MAHOs similar levels of [2H]-incorporation were only achieved for MAHOs 37, 45 and 47 after stirring for 94 minutes (Figure 1). On the other hand MAHOs 23 (phenyl), 40 (2-carboxymethyl ester), 25 (4-methoxyphenyl), 51 (4-chlorophenyl), 52 (4-bromophenyl), 55 (2-



FULL PAPER

bromophenyl) and 56 (4-trifluoromethylphenyl) all afforded reduced levels of [2H]-incorporation i.e. 77% - 94% even after extended reaction times.

11. Further underlying the fundamental difference in reactivity between the oxygen- and sulfur-derived esters was the difference in the observed rates of exchange. In nearly all cases, the aryl thioester motifs underwent hydrogen ® deuterium exchange at significantly faster rates than their oxygen counterparts when compared like-for-like e.g.

A. thiophenol derived 15 faster than phenol derived 23 i.e. k = 195 versus 83. B. 4-bromothiophenol-derived 66 faster

than 4-bromophenol-derived 52 i.e. k = 208 versus 31. C. 4-trifluoromethylthiophenol-derived 68

faster than 4-trifluoromethylphenol-derived 56 i.e. k = 191 versus 67. D. 4-chlorothiophenol 63 faster than 4-

chlorophenol 51 i.e. k = 178 versus 66. E. 4-methoxythiophenol 12 faster than 4-

methoxyphenol 25 i.e. k = 161 versus 60.

Entry Name / [2H]-incorporation Rate

a 4-bromophenol / 14% k = 31

b

c

d

e f

g

h i j

k l

m

n

o

2-bromophenol / 15%

methyl-2-hydroxybenzoate / 24%

4-methoxyphenol / 27%

4-trifluoromethylphenol / 30%

4-chlorophenol / 31%

phenol / 38%


2-cyanophenol / 38%

3-trifluoromethylphenol / 44%

hexafluoroisopropanol / 43%

2,6-dimethylphenol / 46%

2-methoxythiophenol / 46%

4-nitrophenol / 68%

2-chlorothiophenol / 71%

k = 33

k = 53

k = 60

k = 67

k = 66

k = 83

k = 84

k = 85

k = 96

k = 96

k = 101

k = 102

k = 150

k = 156

p

q r

s t

u

v

w

methyl-4-hydroxybenzoate / 72%


2-nitrophenol / 78%

4-cyanophenol / 79%

4-chlorothiophenol / 81%

4-trifluoromethylthiophenol / 88%

thiophenol / 90%

4-bromothiophenol / 96%

k = 157

k = 161

k = 170

k = 171

k = 178

k = 191

k = 195

k = 208

An explanation for the unusual MAHT reactivity

enhancement was required. Initiating a computational study we sought to gain a comprehensive understanding of the reactivity differences. The electronic communication between the aryl ether and aryl thioether functionality and the enol was probed by DFT calculations. First, for the parent MAHO i.e. PhOCH(OH)=CHCO2H, and MAHT i.e. PhSCH(OH)=CHCO2H the E- and Z-enols were compared in two representative conformations: (i) the enol-carboxylic acid section was close to co-planarity and (ii) the plane of the arene was orthogonal (~90°) to the enol-carboxylic acid section.[35]

In view of the similarity of the results obtained with the relatively rapid calculations possible with B3LYP/6-31G(d) or M06-2X/6-31+31G(d,p) compared with MO6-2X/def-QZVP we choose to employ B3LYP/6-31G(d) and M06-2X/6-31+31G(d,p) in the work described here as it would allow us to considerably extend our computational ‘reach’ examining the influence the substituents within the arene have on the conformations and molecular orbitals of the MAHO and MAHTs.[36] Initiating our study on MAHOs twenty-five distinct phenyl MAHO configurations / conformations were identified (see Table S3, SI) the most stable (–648.67222 Ha) retained the phenyl ring at about 92° (dihedral angle –92.14°) relative to the enol-carboxylic acid section which had an almost planar conformation. The enol was in the E-configuration, and the orientation of the carboxylic acid relative to the OPh group was transoid. The H-O-C-C, C-C-C-O and C-C-O-H sections of MAHO 23 and MAHT 15 were “e-z-z”. A frequency calculation, however, indicated the –648.67222 Ha structure was a transition point not a true minimum. Further investigation identified two similar more stable forms (–648.67244 Ha), again corresponding to “E-transoid-e-z-z” structures, but now the phenyl ring was displaced from a perpendicular orientation to –62.3° (Figure 2) and –122.7° (Figure 3).



FULL PAPER

Figure 2. Minimum energy B3LYP/6-31G(d) structures for phenyl ester 23 in –62° dihedral angle form: (A): side view of the enol form of MAHO 23; (B): top view of enol form of MAHO 2

Figure 3. Minimum energy B3LYP/6-31G(d) structures for phenyl ester 23 in –123° dihedral angle form: (A): side view of the enol form of MAHO 23; (B): top view of enol form of MAHO 2

Figure 4. Minimum energy B3LYP/6-31G(d) structures for thioester 15 in –92° dihedral angle form: (A) side view of enol form of MAHT 15; (B): top view of enol form of MAHT 15

These two independently optimized structures had identical energies as well as almost identical bond lengths and angles. The structure at –648.67222 Ha corresponds to the transition between these conformers. As expected, the energy barrier for interconversion was very low i.e. 0.000215 Ha; <1 kJ mol–1.[37] Applying the same computational procedure to the analogous MAHT (15) identified (Table S4, SI) a true minimum

at –971.63728 Ha [B3LYP/6-31G(d) data] as the most stable form. In contrast to the data in Table S5 (SI), the cisoid form was observed, and although a minimum, it had a dihedral angle of –91.67° between PhS and -SC(OH)CHCO2H (Figure 4). An E-transoid MAHT conformation was identified (at –971.63648 Ha) and established to be a transition structure with –91.93° between the PhS and -SC(OH)=CHCO2H. A true E-transoid



FULL PAPER

minimum was found at an almost identical energy with the PhS tilted at –103.3°. As expected, the central bond angle for the MAHTs (~ 103°, see Figure 4) was smaller than the MAHOs (~ 121°, see Figures 2 and 3) and the Ph-S and Ph-O bond lengths (1.80 Å and 1.40 Å respectively) were longer than the PhS-C and PhO-C bond lengths into the enol (1.78 Å and 1.34 Å). The preference for cisoid or transoid forms (for examples, see Figures 2 - 4) in these structures is finely balanced. The calculations were repeated with the highly recommended[38] M06-2X functional[39] using the 6-31+G(d,p) basis set[40] which has been identified[41] as one affording a good balance between performance and computing time for structures with non-covalent or H-bonding interactions. The cisoid conformation was the most stable in both MAHO and MAHT series.

In summary, with such finely balanced conformational preferences identified using B3LYP and MO6-2X and the small energy differences between both the cisoid and transoid series we have established both need to be considered when studying substituent effects on the aromatic ring and interpreting [2H]-incorporation results. In practice, the two conformers easily interconvert and both will be present in solution. It was decided to concentrate initially on the characterization of the more stable E-enols. The “E-cisoid-e-z-z” and “E-transoid-e-z-z” MAHO and MAHT structures (see the Tables S3 and S4 in SI) were now used as the basis to perform calculations on 4-methoxy, 4-chloro and 4-bromo aryl substituted MAHOs and MAHTs. In both the MAHO and MAHT series, cisoid and transoid minimum energy conformations

were identified for the E-enols. Consistently in these structures (Table 7), in which hydrogen-bonding between the enol OH and the C=O of the carboxylic acid was apparent (illustrated in Figures 2 - 4 for the parent phenyl examples), the hydrogen bonding O···H separation was 1.63 - 1.66 Å. The O-H bond of the enol was lengthened to 1.01 - 1.02 Å compared to the carboxylic acid O-H at 0.98 Å and in all cases, the enol and carboxylic acid sections of the structure were essentially coplanar (dihedral angles +0.6 - –0.6°), and the C-C bond lengths within the C=CHCO2H section shorter for C=CH than for CH-CO2H (1.37 - 1.38 Å and 1.43 - 1.44 Å, respectively) but none-the-less these are both significantly shorter than a typical C-C bond, as expected because of the extended conjugation. The molecular orbitals of the gas phase optimised structures were examined from the LUMO down to the lowest p-orbital of the MAHO(T). Figures 5 - 10 illustrate well the flat delocalised p-systems associated with their structures. In general, the patterns of the orbitals were similar for all the MAHOs and MAHTs examined. The p-donation from the oxygen or sulfur atoms of the PhO / PhS groups[42] into the enol is clearly visible in the slightly higher energy E-cisoid HOMO-11[ 43 ] (ester, Figure 7) and E-cisoid HOMO-12 (thioester, Figure 8) molecular orbitals. Similarly, E-transoid HOMO-8 and E-transoid HOMO-10 show the delocalisation between the heteroatom link and the arene (Figure 9), especially in the thioester case. In Figures 5B, 7B and 9B, the tilted geometry of the MAHO distorts these orbitals.

Figure 5. Low lying E-cisoid and E-transoid MOs in the ester structures with p-overlap across the enol and carboxylic acid (A: structure cisoid-23; B: structure transoid-23



FULL PAPER

Figure 6. Low lying E-cisoid and E-transoid MOs in the thioester structures with p-overlap across the enol and carboxylic acid (A: structure cisoid-15; B: structure transoid-15)

Figure7.Typicalexamplesofp-delocalisationintheparentMAHO23withp-overlapacrosstheenolandcarboxylicacid(A:structurecisoid-23;B:structuretransoid-23)

Figure 8. Typical examples of p-delocalisation in the parent MAHT 15 with p-overlap across the enol and carboxylic acid (A: structure cisoid-15; B: structure transoid-15).

Figure 9. Orbitals illustrating p-overlap between the heteroatom and the arene (A: structure cisoid-15; B: structure transoid-15)

Figure 10. Orbitals illustrating p-overlap between the heteroatom and the arene (A: structure cisoid-15; B: structure transoid-15)



FULL PAPER

Figure 11. Orbitals illustrating p-overall between the heteroatom and the arene (A: structure cisoid-15; B: structure transoid-15) These archetypal p-symmetry orbitals are observed

across the whole series of MAHO and MAHT structures (see SI) with only minor variations in the sequence of relative energies. A significant difference between MAHOs and MAHTs, however, was identified from the presence of an additional frontier orbital (HOMO-3) in the thioesters (see Table 8, and Figure 11) which is delocalized between sulfur atom and the enol. In other aspects, the geometries and molecular orbitals of the MAHOs and MAHTs are very similar and consistently showing the same key features.[44] The significant HOMO-3 MAHT orbital, however, which was consistently present in all MAHT conformers examined in our study was not present in the MAHOs and provides for increased p-overlap in the MAHT series. This may account for the much easier enolisation and much faster deuterium incorporation observed in the NMR experiments for the parent phenol and thiophenol MAHO and MAHT structures.

Conclusions

In summary, we detail the synthesis of 39, mostly unknown, structure- and function-diverse MAHOs and MAHTs.

All are easily prepared using a quick and efficient protocol. Whilst the yields in most cases are good, the inclusion of a strong EWG i.e. nitro or ester at the ortho-position on the aryl ring of the phenol or thiophenol starting materials affords the products, but with reduced yields. This is attributed to the formation of a strong intramolecular hydrogen-bond (cf. RAHB) between the hydroxyl or thiol group and the ortho-EWG. Overall, the formation of a resonance-assisted hydrogen bond and the inductive effect reduces nucleophilicity and reactivity towards the malonyl mono-chloride. Whilst probing the reactivity of the MAHO and MAHTs, as measured by hydrogen ® deuterium exchange, we identified an unexpectedly rapid enolisation in the MAHT series. We ascribed the increased reactivity to the presence of an additional sulfur-centred frontier orbital which increases p-delocalisation between the heteroatom and the enol. The exceptional ease of enolisation in the MAHT series suggests these nucleophiles should offer new prospects for the efficient C-C bond formation under mild conditions. Ultimately, our work will help develop a better understanding of how Nature employs MAHTs for metabolite biosynthesis and with the two slightly different protocols now well established, we have the capability to synthesize either electron-rich, or electron-poor or sterically hindered aryl derived MAHOs or MAHTs.



FULL PAPER

Acknowledgements

The authors acknowledge the EPSRC, UEA, and INTERREG IVA (IS:CE chem project 4061) for financial support, the UEA high-performance cluster “Grace” for computational resources,

[1] a)G.Zhou,D.Lim,D.M.Coltart,Org.Lett.2008,10,3809–3812;b)A.I.Scott,C.J.Wiesner,S.Yoo,S.K.Chung,J.Am.Chem.Soc.1975,97,6277-6278;c)Y.Kobuke,J.-IYoshida,TetrahedronLett.1978,19,367-370;d)M.W.Rathke,P.J.Cowan,J.Org.Chem.1985,50,2622-2624;e)A.L.Gutman,A. Boltanski,Tetrahedron Lett.1985,26, 1573 - 1576; f) R. J. Clay, T. A.Collom,G.L.Karrick,J.Wemple,Synthesis1993,290-292;g)X.Wang,W.T.Monte,J.J.Napier,A.Ghannam,TetrahedronLett.1994,35,9323-9326;h)Y.Ryu,A.I.Scott,TetrahedronLett.2003,44,7499-7502;i)N.Sakai,N.Sorde,S.Matile,Molecules,2001,6,845-851.

[2] a)K.C.Fortner,M.D.Shair,J.Am.Chem.Soc.2005,127,7284–7285;b)D. J. Schipper, S.Rousseaux,K. Fagnou,Angew.Chem. Int.Ed.2009,48,8243–8246;c)S.Orlandi,M.Benaglia,F.Cozzi,TetrahedronLett.2004,45,1747-1749;d)G.Lalic,A.D.Aloise,M.D.Shair,J.Am.Chem.Soc.2003,125,2852-2853;e)N.Blaquiere,D.G.Shore,S.Rousseaux,K.Fagnou,J.Org.Chem.2009,74,6190–6198;f)J.Baudoux,P.Lefevre,M.-C.Lasne,J.Rouden,GreenChem.2010,12,252-259;g)K.Rohr,R.Mahrwald,Org.Lett.2011, 13, 1878 - 1880; h) X.-J. Li, H.-Y. Xiong,M.-Q.Hua, J.Nie, Y.Zheng,J.-A.Ma,TetrahedronLett.2012,53,2117-2120.

[3] a)L.Yin,M.Kanai,M.Shibasaki,J.Am.Chem.Soc.2009,131,9610–9611;b) Y. Umezawa, H. Morishima, S.-I. Saito, T. Takita, H. Umezawa, S.Kobayashi,M.Otsuka,M.Narita,M.Ohno, J.Am.Chem.Soc.1980,102,6631-6633;c)M.Otsuka,M.Narita,M.Yoshida,S.Kobayashi,M.Ohno,Y.Umezawa,H.Morishila,S.Saito,T.Takita,H.Umezawa,Chem.Pharm.Bull.1985,33,520-526;d)A.Ricci,D.Pettersen,L.Bernardi,F.Fini,M.-F.Fochi,R.PerezHerrera,V.Sgarzania,Adv.Synth.Catal.2007,349,1037-1040;e)Y.Pan,C.Kee,Z.Jiang,T.Ma,Y.Zhao,Y.Yang,H.Xue,C.-HTan,Chem.Eur.J.2011,17,8363-8370;f)L.Yin,M.Kanai,M.Shibasaki,M.Tetrahedron2012,68,3497-3506.

[4] a)J.Lubkoll,H.Wennemers,AngewChem.Int.Ed.2007,46,6841–6844;b)J.E.McMurry,W.A.Andrus,J.H.Musser,Synth.Commun.1978,8,53-57;c)M.Banwell,D.Hockless,B.Jarrott,B.Kelly,A.Knill,R.Longmore,G.Simpson,J.Chem.Soc.,PerkinTrans.12000,3555-3558;d)M.MakotoFurutachi, S.Mouri, S.Matsunaga,M.Shibasaki,Chem.Asian J.2010,5,2351-2354;e)H.Y.Bae,S.Some,J.H.Lee,K.-Y.Kim,M.J.Song,S.Lee,Y.J.Zhang,C.E.Song,Adv.Synth.Catal.2011,353,3196-3202.

[5] N.Blaquiere,D.G.Shore,S.Rousseaux,K.Fagnou,J.Org.Chem.2009,74,6190–6198

[6] a)K.I.Arnstadt,G.Schindlbeck,F.Lynen,Eur.J.Biochem.1975,55,561–571;b)M.J.S.Dewar,K.M.Dieter,Biochemistry1988,27,3302-3308.

[7] K.C.Fortner,M.D.Shair,J.Am.Chem.Soc.2007,129,1032-1033 [8] a) C. Khosla, Chem. Rev. 1997, 97, 2577 – 2590; b) E. S. Sattely, M. A.

Fischbach,C.T.Walsh,Nat.Prod.Rep.2008,25,757–793;c)L.Van,G.Steven,B.Shen,CurrentOpinioninDrugDiscovery&Development2008,11,186–195;d)R.J.Cox,Org.Biomol.Chem.2007,5,2010–2026;e)M.A. Fischbach, C. T. Walsh, Chem. Rev. 2006, 106, 3468 – 3496; f) K. J.Weissman,P.F.Leadlay,NatureReviewsMicrobiology2005,3,925–936;g)D.E.Cane,C.T.Walsh,C.Khosla,Science1998,282,63-68.

[9] a) E. M. Musiol, T. Weber,MedChemComm 2012, 3, 871 – 886; b) F.Ishikawa,R.W.Haushalter,M.D.Burkart,J.Am.Chem.Soc.2012,134,769–772;c)H.Morita,M.Yamashita,S.-P.Shi,T.Wakimoto,S.Kondo,R.Kato,S.Sugio,T.Kohno,I.Abe,PNAS2011,108,13504-13509;d)J.J.Gehret,L.Gu,W.H.Gerwick,P.Wipf,D.H.Sherman,J.L.Smith,J.Bio.Chem.2011,286,14445–14454;e)D.H.Kwan,P.F.Leadlay,ACSChem.Biol.2010,5,829-838.

[10] A)Z.-L.Wang,Adv.Synth.Catal.2013,355,2745–2755;b)S.Nakamura,Org.Biomol.Chem.2014,12,394-405;c)Y.Pan,C.-H.Tan,Synthesis,2011,2044–2053;d)H.Y.Bae, J.H.Sim,J.-W.Lee,B.List,C.E.Song,AngewChem.Int.Ed.2013,52,12143–12147;e)H.-N.Yuan,S.Wang,J.Nie,Q.Yao,J.-A.Ma,AngewChem.Int.Ed.2013,52,3869–3873;f)N.Hara,S.Nakamura,M. Sano, R. Tamura, Y. Funahashi, N. Shibata, Chem. Eur. J.2012,18,9276–9280;g)S.Nakamura,M.Sano,A.Toda,D.Nakane,H.Masuda,Chem.Eur.J.2015,21,3929–3932;h)H.Y.Bae,Synlett,2015,26,705–706;i)H.Y.Bae,C.E.Song,Bull.Korean.ChemSoc.,2014,35,1590–1600;j)L.Bernardi,A.Ricci,M.C.Franchini,Curr.Org.Chem.,2011,15,2210–2226;k)Y.Cotelle,V.Lebrun,N.Sakai,T.R.Ward,S.Matile,ACSCentralScience,2016,2,388–393;l)J.Saadi,H.Wennemers,Nat.Chem.

and the EPSRC Mass Spectrometry Service for generating HRMS data.

Keywords: thioester • isotope exchange • DFT • reaction rate • hydrogen bond • p-delocalisation

2016,8,276–280;m)Q.Ren,T.Gao,W.Li,H.Wan,P.Yimin,S.Yanhong,S.Sun,L.Hu,M.Wu,H.Guo,NewJ.Chem.,2015,39,5100–5103.

[11] N.Utsumi,S.Kitagaki,C.F.Barbas,III,Org.Lett.2008,10,3405-3408.[12] Y.Pan,C.W.Kee,Z.Jiang,T.Ma,Y.Zhao,Y.Yang,H.Xue,C.-H.Tan,Chem.

Eur.J.2011,17,8363-8370.[13] H.Y.Bae,S.Some,J.H.Lee,J.-Y.Kim,M.J.Song,M.J.Lee,Z.Sungyul,C.E.

Song,Adv.Synth.Catal.2011,353,3196-3202.[14] S.Nakamura,M.Sano,A.Toda,D.Nakane,H.Masuda,Chem.Eur.J.2015,

21,3929-3932.[15] A.Ricci,D.Pettersen,L.Bernardi,F.Fini,M.Fochi,R.P.Herra,V.Sgarzani,

Adv.Syn.Cat.2007,349,1037–1040.[16] A.S.Ivanov,Chem.Soc.Rev.2008,37,789-811.[17] Eds.: W. L. F. Armarego and D. D. Perrin, Purification of Laboratory

Chemicals6thEdition,Butterworth-Heinemann,Oxford,2009.[18] S.-J.Park,J.-C.Lee,K.I.Lee,Bull.KoreanChem.Soc.2007,28,1203-1205.[19] T.Mase,H.Arima,K.Tomioka,T.Yamada,K.Murase,Eur.J.Med.Chem.

1988,23,335–339.[20] T.Imamoto,M.Kodera,M.Yokoyama,Bull.Chem.Soc.Jpn.1982,55,2303-

2304.[21] a) S. P. Bew, G. R. Stephenson, J. Rouden, L. A. Martinez-Lozano, H. Seylani, Org. Lett. 2013, 15, 3805-3807; b) F. N. Miros, Y. Zhao, G. Sargsyan,M.Pupier,C.Besnard,C.Beuchat,J.Mareda,N.Sakai,S.Matile, Chem.Eur.J.2016,22,2648-2657 [22] Monomalonyl chloride is commercially available from Akos

(http://www.akosgmbh.eu/), Aurora Fine Chemicals(http://www.aurorafinechemicals.com/) and FCH Group(http://www.fchgroup.net).

[23] a)M.S.Singh,S.Chowdhury,RSCAdvances2012,2,4547–4592;b)J.G.Hernandez, E. Juaristi, Chem. Comm. 2012, 48, 5396 – 5409; c) H. M.Marvaniya,K.N.Modi,D.J.Sen,Int.J.DrugDev.&Res.2011,3,34–43;d)V. P. Shevchenko, I. Y. Nagaev, N. F.Myasoedov, J. Labelled Comp. andRadiopharmaceuticals 2010, 53, 693 – 703; e) W. C. Shearouse, D. C.Waddell,J.Mack,CurrentOpinioninDrugDiscovery&Development2009,12,772–783;f)M.A.P.Martins,C.P.Frizzo,D.N.Moreira,L.Buriol,P.Machado,Chem.Rev.2009,109,4140-4182.

[24] A.Heintz,S.Kaptenia,S.P.Verekin,J.Phys.Chem.A.2007,111,6552-6562[25] K.B.Borisenko,C.W.Bock, I.Hargittai, J.Phys.Chem.,1994,98,1442–

1448.[26] F.Iwasaki,YKawano,Acta.Cryst.B1978,38,1286-1288.[27] J.Zheng,K.Kwak,X.Chen, J.B.Asbury,Ma.D.Fayer, J.Am.Chem.Soc.

2006,128,2977–2987. [28] a)B.Rigo,D.Fasseur,P.Cauliez,D.Couturier,TetrahedronLett,1989,30,

3073–3076;b)X.-J.Li,H.-Y.Xiong,M.-Q.Hua,J.Nie,Y.Zheng,J.-A.Ma,TetrahedronLett.2012,53,2117-2120.

[29] M. H. Abraham, R. J. Abraham, A. E. Aliev, C. F. Tormena, Physical ChemistryChemicalPhysics2015,17,25151-25159.[30] H.Heydari,HRaissi,F.Mollania,Struct.Chem.2015,26,971–987.[31] T.Kobayashi,A.Yamashita,Y.Furuya,R.Horie,M.HirotaBull.Chem.Soc.

Jpn.1972,45,1493-1498.[32] a) G. Hilt,F. Pünner,J.Möbus, V. Naseri,M. A. Bohn,Eur. J. Org. Chem.

2011, 30, 5962 - 5966; b) D. Sheppard, D.-W. Li, R. Brüschweiler, V.Tugarinov, J.Am.Chem.Soc.2009,131,15853-15865;c)N.F.Breen,T.Weidner, K. Li,D.G.Castner,G. P.Drobny, J. Am.Chem. Soc.2009,131,14148-14148;d)Y.Liu,R.Warmuth,Org.Lett.2007,9,2883-2886;e)B.Reif,Y.Xue,V.Agarwal,M.S.Pavlova,M.Hologne,A.Diehl,Y.E.Ryabov,N.R.Skrynnikov,J.Am.Chem.Soc.2006,128,12354-12355;f)Y.-L.ZhaoK.N.Houk,R.Dalit,A.Scarso, J.Rebek, Jr., J.Am.Chem.Soc.2004,126,11428-11429;g)S.Bolvig,E.P.Hansen,Mag.Res.Chem.1996,34,467-478; h) K. L. Servis,R. L. Domenick, J. Am. Chem. Soc.1985,107, 7186 -7189; i) D. Bal, A. Kraska-Dziadecka, A. Gryff-Keller, J. Org. Chem.2009,74,8604 - 8608; j) P. E.Hansen, J. LabelledCompds. Radiopharm.,2007,50,967-981;k)R.M.Claramunt,C.Lopez,M.D.SantaMaria,D.Sanz,J.Elguero,ProgNuc.Mag.Res.Spect.2006,49,169 -206; l)S.Bolvig,F.Duus,E.P.Hansen,Erik,Mag.Res.Chem.1998,36,315-384.

[33] a)H.R.Lu,M.-G.Gu,Q.Huang,J.J.Huang,W.-Y.Lu,H.Lu,Q.-S.Huang,Antimicrobial Agents and Chemotherapy, 2013,57,1872 - 1881; b) S.Mehmood, C. Domene, E. Forest, J.-M. Jault, PNAS, 2012,109,10832 -



FULL PAPER

10836;c)K.-Y.Chung,S.G.F.Rasmussen,T.Liu,S.Li,B.T.DeVree,P.-SChae, D. Calinski, B. K. Kobilka, V. L.Woods Jr.; R. K. Sunahara,Nature,2011,477,611-615;d)F.Ito,S.Ando,M.Luchi,T.Ukari,M.Takasaki,K.Yamaguchi, Tetrahedron,2011,67,8009 - 8013; e) D. Houde, S. A.Berkowitz,PharmaceuticalTechnology,2010,s12,s14;f)E.V.Kudrik,A.B.Sorokin,Chem. Eur. J, 2008,14,7123-7126; g) M. Chen, K. D. Cook, I.Kheterpal,R.Wetzel,J.Am.Soc.MassSpec.2007,18,208-217;h)S.Y.Dai,M.C.Fitzgerald,Biochemistry,2006,45,12890-12897;i)T.J.D.Jorgensen,H.Grdsvoll,M.Ploug,P.Roepstorff, J.Am.Chem.Soc.2005,127,2785 -2793;j)V.L.Woods,Jr.;Y.Hamuro,J.CellBiochem.,2001,37,89-98;k)I.D.Figueroa,D.H.Russell,J.Am.Soc.MassSpect.,1999,10,719.

[34] S.Taoka,R.Padmakumar,C.B.Grissom,R.Banerjee,Bioelectromagnetics1997,18,506-513.

[35] a)EachwasoptimizedusingGausian09 [35b]withtheB3LYPfunctional[35c]and the 6-31G(d) basis set [35d] in gas phase calculations. The choice offunctional and basis set for the first stage of our work has been firmlyestablished i.e. benchmarked in our preliminary calculations[21] with theparent thiophenyl-derivedMAHT (15) inwhich very similar resultswereobtainedwith bothB3LYP andMO6-2X functionals and the 6-31G(d), 6-31+G(d,p)anddef-QZVPbasissets;b)Gaussian09,Revision–M.J.Frisch,G.W.Trucks,H.B.Schlegel,G.E.Scuseria,M.A.Robb,J.R.Cheeseman,G.Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M.Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L.Sonnenberg,M. Hada,M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa,M.Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A.MontgomeryJr.,J.E.Peralta,F.Ogliaro,M.Bearpark,J.J.Heyd,E.Brothers,K.N.Kudin,V.N.Staroverov,R.Kobayashi,J.Normand,K.Raghavachari,A.Rendell,J.C.Burant,S.S.Iyengar,J.Tomasi,M.Cossi,N.Rega,N.J.Millam,M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R.Gomperts,R.E.Stratmann,O.Yazyev,A.J.Austin,R.Cammi,C.Pomelli,J.W.Ochterski,R.L.Martin,K.Morokuma,V.G.Zakrzewski,G.A.Voth,P.Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas, J. B.Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox,Gaussian 09, revision A.1;Gaussian,Inc.:Wallingford,CT,2009;c)A.D.Becke,J.Chem.Phys.1993,98, 5648−5652; d) C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37,785−789;e)R.Ditchfield,W.J.Hehre,J.A.Pople,J.Chem.Phys.1971,54,724-726;f)TheMO6-2X/def-QZVPleveloftheoryhasbeenvalidatedbyFogueri et al. in a full benchmarked study incorporating 52 melatoninconformersseeU.M.Fogueri,S.Kozuch,A.Karton,J.M.L.Martin,J.Phys.Chem. A 2013, 117, 2269 - 2277. A computational study on dimethylmonothiocarbonatebyVedovaetal.invokedananomericandmesomericeffectsandnaturalbondordertoexplaintheconformationalpreferences,seeM.F.Erben,R.Boese,C.O.D.Vedova,H.Oberhammer,H.WillnerJ.Org.Chem.2006,71,616-622.

[36] Of relevance to our computational work onMAHOs andMAHTs was asuccessfulstudyoncis-transamidebondrotamersinpeptoidsthisincludedexampleswith thioamidebonds see J. S. Laursen, J. Engel-Andreasen,P.Fristrup,P.Harris,C.A.Olsen,J.Am.Chem.Soc.2013,135,2835−2844.

[37] VerysimilarresultsfortheMAHOserieswereobtainedfromcalculationsusingMO6-2X/6-31+G(d,p) with energies ranging from –648.33202 to –648.31256Ha[cf.–648.67243to–648.64890Ha,B3LYP/6-31G(d)data].

[38] E.G.Hohenstein,S.T.Chill,C.D.Sherrill,J.Chem.TheoryComput.2008,4,1996–2000.

[39] Y.Zhao,D.G.Truhlar,Theor.Chem.Acc.2008,120,215–241.

[40] I.D.Mackie,G.A.DiLabio,J.Phys.Chem.A2008,112,10968–10976.[41] S.O.NilssonLill,J.Phys.Chem.A2009,113,10321–10326.[42] a)S.Nagata,T.Yamabe,K.FukuiJ.Phys.Chem.1975,79,2335–2340;b)M.

Geronés,A.J.Downs,M.F.Erben,M.Ge,R.M.Romano,L.Yao,C.O.D.Védova,J.Phys.Chem.A2008,112,5947–5953;c)A.Rai-Chaudhuri,W.ShongChin,D.Kaur,C.Y.Mok,HsingH.HuangJ.Chem.Soc.,PerkinTrans.2,1993,1249-1250;d)N.Y.Dugarte,M.F.Erben,R.M.Romano,M.-F.Ge,Y.LiandC.O.D.VédovaJ.Phys.Chem.A,2010,114,9462–9470.

[43] The notation HOMO–n indicates the sequence of molecular orbitals indescending order of energy, starting from the HOMO, then HOMO–1,HOMO–2,etc.

[44] Forexampleseea)Figure5(MAHO)andFigure6(MAHT)withp-overlapacross theS-enol-CO2Hsectionof thestructure;b)Figure7 (MAHO)andFigure8 (MAHT)p-overlapwithoneadditionalverticalnode; c)Figure5(MAHO)andFigure6(MAHT)additionaldelocalisationofelectrondensitythroughtheheteroatom.

[45] G.E.WilsonJr.,A.Hess,J.Org.Chem.1974,39,3170-3170.

MAHTS are the best. Three comprehensive studies on the synthesis, reactivity ([2H]-exchange) and conformational properties of aryl (thio)ester-derived MAHOs and MAHTs are reported. Using ‘off the shelf’ starting materials and a solvent-free protocol 39 structure and function diverse MAHOs and MAHTs are readily generated. These are the fastest, easiest and most environmentally friendly routes to MAHO(T)s. The synthetically valuable ease of enolisation of the MAHTs is accounted for by the computationally established presence of an additional frontier orbital unique to the MAHT series which provides enhanced p-delocalisation between the sulfur and the enol.



ac ce pt ed articl e · however, 1 equivalent of malonic acid afforded a 40% yield, a 44% drop when...

Documents